1. Example
   1. From raw text to gorgeous HTML
   2. General idea behind RMarkdown
2. 3 Component chunks
   1. Metadata
   2. Text (Markdown)
   3. Code chunks
3. Additional perks
   1. pandoc
   2. BiBTeX and reference management
   3. CSS customization
   4. To the infinity and beyond

• This whole presentation (ioslides) was made in RMarkdown.

regression <- linregression_traditional(data,X,Y,carnegie)

• We observe an \(R^2\) value of 0.9072477, which is quite high
• For every research/instructional employee, we have 3 staff employees!

Work flow

• A single plain text document to produce
  • High quality reports: PDF, HTML, Word
  • High quality presentations: Slidy, ioslides, beamer, PowerPoint
  • To the infinity and beyond: flexboards, documentation,books, thesis, etc.
• Integrate multiple languages: R, python, bash, julia, etc.
• The basics come bundled with

install.packages('rmarkdown')

• This whole presentation is based on Xie Yihui’s extensive documentation on RMarkdown and its spin-offs (Xie, Allaire and Grolemund, 2018)

• Metadata indicated at the top sets the general behavior of the document
• As in pandoc, use YAML (established format)
• Tabs matter

---
title: Walking through RMarkdown
author: 
   - Erik Amézquita
   - Computational Math, Science & Engineering
   - Michigan State University
   - " "
   - amezqui3@msu.edu
   - https://egr.msu.edu/~amezqui3
date: May 18th, 2020
bibliography: biblio.bib
csl: harvard-cite-them-right.csl
output: 
  ioslides_presentation:
    smaller: True
    css: styles.css
    logo: "figs/cmse_logo_transparent.png"
---

• Check the cheatsheet for all the possible options.

1. Markdown is a markup language: f^orma_t text with some quick rules.
   1. Look at the previous reference or at pandoc’s documentation.
   2. No more need to click here and there
      • No need to use shortcuts

## 3 Components: Text = pandoc's markdown

1. **Markdown is** a _markup language_: f^o^rma~t~ text ~~with some~~ `quick rules`.
    1. Look at the [previous reference](https://rstudio.com/wp-content/uploads/2015/03/rmarkdown-reference.pdf) or at [pandoc's documentation](https://pandoc.org/MANUAL.html#pandocs-markdown). 
    2. No more need to click here and there
        + No need to use shortcuts

Table Header  | Second Header
------------- | -------------
Tables        |  are
literally     |  drawn

• Mathematical typeseting \(\displaystyle \int_\alpha^\infty \frac{dt}{t\log t} = \log(\log t)\bigg|_\alpha^\infty = \infty\) is supported by \(\LaTeX\).
• Put the \(\LaTeX\) code in between $ symbols as usual

$\displaystyle \int_{\alpha}^{\infty} \frac{dt}{t\log t} 
= \log(\log t)\bigg|_{\alpha}^{\infty} = \infty$

• Superscripts with ^: \(2^{10}=1024\) is $2^{10}=1024$
• Subscripts with _: \(t_{10}\) is $t_{10}$
• A quick list of common symbols can be found at the end of this document
• As a personal note, \(\LaTeX\) typesets the most beautiful documents you’ll find out there. It is worth to look into more detail. This is one of the many resources out there.

knitr::include_graphics(c('figs/S017_L3_26.png', 'figs/S017_L4_01.png', 'figs/seed2_X.gif'))

caption

• More flexibility. Take advantage of the full power behind RMarkdown.

• Pieces of a usual coding script
• Wrangle data, crunch numbers, plot results, showcase tables, etc
• These are delimited with ````` with a few specifications
  • Specify the programing language
  • Label (optional)
  • Diverse options: show code? Show result? Center the plot? Size of the graph?

     {r label, options}
     some R code

• In RStudio, use the Ctrl + Alt + I shortcut to make a new code chunk (Cmd + Option + I on macOS).
• Check the cheatsheet for all the possible code chunk options.

• Some default options can be defined globally at the beginning via opts_chunk$set()

# do not show code unless stated otherwise
knitr::opts_chunk$set(echo = FALSE) 
# do not include the chunk in the final version
knitr::opts_chunk$set(include = FALSE) 
# point RMarkdown to the right python engine
knitr::opts_chunk$set(engine.path = list(python = '/usr/bin/python3'))

• Solve an ODE with python3 via scipy.integrate.odeint

import matplotlib.pyplot as plt
import numpy as np
from scipy.integrate import odeint

def f(y, t, k=0.2):
  r=(1+2*np.sin(t)-np.cos(t))/5
  return r*y-k

t  = np.linspace(0, 20., 1000)

for i in range(6):
    y0=0.1*i+0.5
    soln = odeint(f, y0, t)
    plt.plot(t,soln, label="y0="+str(y0))

plt.xlabel("time")
plt.ylabel("population")
plt.title("Population Dynamics")
plt.legend()
plt.grid()

• ODE \(\frac{dy}{dt} = r(t)y-k, \quad y(0)=y_0\)
• Growth rate \(r(t) = \frac{1}{5}(1+2\sin t - \cos t)\)
• Constant predation rate \(k=\frac{1}{5}\)
• Explore which initial population value \(y_0\) is critical to avoid extintion.

plt.show()

echo "The following RMarkdowns were found here"
for file in *.Rmd
do
  echo $file" --->  ""${file%.*}"".html"
done

## The following RMarkdowns were found here
## doctoral_universities.Rmd --->  doctoral_universities.html
## intro_to_rmarkdown.Rmd --->  intro_to_rmarkdown.html

• caveat: unlike r, we won’t be able to save results from bash for other chunks!
• however, python carries its variables from chunk to chunk like r
• nonetheless, python and r live in separate environments
• check here for more details on how exchange information between r and python via the reticulate library in r.

rmarkdown::pandoc_convert('day2_summary.md', output='day2_summary.pdf')

Taking easy notes

• From markdown, TeX, word, odt, rtf, html, ipynb, ebook, etc.
• To PDF, beamer, markdown, TeX, word, odt, rtf, html, etc.
• Go to the official pandoc documentation for more details.